A eTA PALA EO;\,TOLOGICA POLONICA vet. I 1 956 No . 2

JULIAN KULCZYCKI

ON THE PARASPHENOID OF THE BRACHYTHORACI

A bst ract . - A description is given of the Brachythoraci parasphenoid as compared w ith th e corresponding structure dm Ao ant~idiiJalIl!S and some Crossopterygnans (Po­ rolsp idae jvand ',viloh dhe Acanthodierr -anter-ior-basai.

INTRODUCTION

The fi rst desc rip tion 0.£ the parasphenoid 0.£ the Arthrodires was made by E. S. Hills (1936) upon discovering inthe endocranial floor of 1Jhe Coc­ costeus (= Buchanosteus) osseus the presenceof 'a "thiokened part com­ pos ed orf solid bone' with a som ewhat complicated structure. A dermal element, coating the very same region of eIIldlOcT'ani~m of I Kujdanovia spis buczacziensis but showing a much simpler structure, was des cr ib ed by E. A. Stensio (1942). . In an. addendum to his next paper on the cramum and cervical jodnt of cth e Dolichothoraci (Acanthaspida) (1945), Stensio is 10010 'the conclu­ sion thetthe dermal element, decribed by him and by Hidlsvconstdtutes the p arasphen oid. In the same paper Stensio writes thaJt in certain Bra­ chythoracid Arthrodires from Wildungen (Erromenosteus, Trematosteus, Brachyosteus) he had ascer-tained ,the presence of the parasphenodd, sha­ ped as .a powerful bone, extendin g, as it does in the Crossopteryglans and most of the Paiaeonisciformes, backwards to underneath the tran­ sition al part between the orbito-temporal and the otic regions of the endocr anium.E. A. Stensio believes fu:at th~s element' is provided with a paired processus ascendens, In their paper on the anterior superogn athal of Gorgonichthys, D. H. Dunkle and P. A. Bungart (1946) published a dJr.aiWdalg representing the parasphenoid of Dinichthy3 terreUi amd Heintzichthys gouldii mentio­ ning in the text fhat the SUiPeI':ognartlha:l elements are iJnJtimaJtely associa­ ted with a parasphenoid-ljke plate which is apparently to be homologised with the arsterior-basal element of the Acanthodians. The most recent reference to this element, here under consideration, is to be found iJn a paper by E. Jarvik (1954). When speaking of his hypothesis concerning the developmenr of parasphenoid in fishes, that 104 J UI.IAN KULCZYCKI author expr esses an opinion that in the Arthrodires the home, as m en­ tion ed above, r ep resents a primitive typ e of structure, apparently w ith out ascending p rocesses. Basing h is state men t on an unpublished paper by StOOo9i6 h e also alleges that th e developmen t of the p arasphenoid in Arthrodires is characterised b y disintegrati on 'an d r educ tion. Kujdanoviaspis buczacziensis and Buchanosteus osseus are the only r epresentatives of 'the Anthrodires w hose parasphenoids have been re­ ported upon at any lengrth. Since, as sh own by Stensio, both, th e latter of t hese forms, as well as Kujdanoviaspis, belong to. the Arctolep id a, no de tasled d escription of the par asphenoid of Brachy thorac i is thus far available. There a re three sp ecimen s of these bones in the fossil m ater ial dep o­ sited at the Muzeum Ziemi (Museum of the Ear th) in Wa rszawa a nd collected by l ate Mrs. Z. Gorizdro-Kulczycka in ,'fue Upper Devonian of Kieke. Two of them w ere discovered in association with other skull fragments belonging to DinichthYJ d . pu stulosus , and to 'the represen ­ tative of a new genus of Brachythoraci, probably closely r elated to Pholidosteus. A descr ip tion of these fragments will be given, in a forth­ coming paper. No fossil r em ains were found associat ed w ith the th ird specimen. Land d: ts generic and family appur tenance, cannot, therefore, be established. Never-theless its close resemblance to the parasphenoid of Dinich thys terrelli and Heintzichthys gouldii as fi­ gur ed in the drawing by Dunkle and Bun gart , testifies to its refe­ l'ab ility to the Brachyithorac i. This specim en is in the m ost complete state of pr eservation of the three at the writer's disposal. T hrough treat­ ment in acetic acid som e of its very d elicate structur al details h ave been disclosed and it is on this specimen that the writer has based h is p r esent description of the parasphenoid of Br achy thoraci. The writer's thanks are due to Professor D .W. Obrutseh ew of Moscaw a nd to Dr. D. H. Denison of Chicago for the valuable h ints received. from them by way of correspondance wh ile he was workin g on h is description of Devoniari fishes fromthe Holy Cross Mts. He is also p articularly m­ debted to Professor W. Gross of the Berlin Humbold t University for makirng accessible to him the collection of fosssl fishes

DESCRIP T IO N The parasphenoid of the Brachyfhoraci displays a r ather complicated structure. Two parts may be differentiated in t his h one, n amely the anterior (pa) tin the form of a s trong plate w ith a sub-cordate outline, ox H IE PARASP HDIOID OF THE BRACHYTHORA CI 105

terminating in a sharp apex projected anteraoriy, and the posterior (pp.) tran sversally elongated , thick around the midpoirit and thin around the edges . Tr an sversally the an t erior cor dat e p ar t (pa.) .is somewhat concave on the ventral side, and convex dorsally . At the sa m e tim e its lateral edges rather mm up in the dorsal direction, p anticular'ly so in rts broad poste­ rior region . Longitud in ally t he ante rior cordate p art bends gradually dorsad as it approaches itsantenior limit. On the ventral side, along the mid -line, a longitudinal crest (cr. m.) occurs in some forms (though ab­ sent in the specim en incertae fam iliae, showing the best state of preser­ vation ), more or less sharply edged , w ith h eight in cr easing anterior ly. Its basal wid th increases in the same direction. The crest is par ticular.l y d istinct in the parasphenoid of Dinichthys cf. pustulosus where it r eaches tcfhe anterior end of the elemerrt iand rt er rninates 'in a sharpened pro­ cess. In t his form the longutud in al crest p asses posteriorl y into a rather small fl at area (a. tub.) spotted b y minute tubercles identical to those found on the external surface of t he dermal bones of the skull. A longi­ tu dinal groove r uns dorsally along the mid-line of the anterior cordat e par t of the paeasphenoid (sm.). The free bone edge is grooved and ser­ rated. The pos terior p art (pp .) is unfortunately in 'a state of incomplete pre­ se rvati on. The lacking d etails are the more delicate marginal parts, in the form of thin lat eral lamellar processes (prl. v .), situated ventrally, w hose ou tl in e h as b een figured b y Dunkle and Bungart (1946) In their drawing of t he parasphenoid of Dinich th ys terr eHiand Heintzichthy s gouldii. On the base of the cited drawing it may be :iJnf,eu-r ed, that these processes are very well d evelop ed in Dinichthys terrelli, w hile irn Hein­ t Z i c h t~ys gouldii they ar e rather rudim entary. The p osterior part of the parasp henoid of Br achythor aci tran sversally somewh at con cave, prese nts a perfectly smooth and even ventral surface, Jack ing an y d etails. It is separ ated from the anterior corda te part by a di­ stin ct transverse groove (s. tr.) wh ose course across 'fu e element is tr aced at an angl e with anteriorly p oint ed ap ex . Medially the gr oove is parti­ tion ed by a bon e lamella pierced by a large foramen communicating w ith the left and right portions of the groove. 'Dhis gives the semblance of a delicate bone-bridge spanning the oern!uml region of the groove (p .). The ven t ral side olf iJhe bridge p aJSiS€ S wirth mo signs of demarcation to the ven tral surface of the anterior corda te part, while from the posterior part ,it is se parated by a narrow groove ccmmun icating a t both ends with the tr ansverse groove w hich transacts the whole element. Close to the bone m argin a minor groove branch es off on both sides into the drans­ ve rse groove (pl, II, fig. 7) and traverses the lateral surface running 00- 106 J ULI AN KULCZYCKI teriorly to the dorsal side of thearsteriorcordate pa.rt of the element, where it soon disa pp ears. In the mid-part of the groove, on both sides of the m edian bridge, there are two sym metrically arranged, faiTl y large foramina (c. aci.) separat ed by a bone lamella (pl. I, fig. F ; pl. II . fig. 6), vertically piercing 1Jhe parasphenoidand communicating w ilith a ba­ sin-like depression situated centraily on the d or sal side of th e elem ent. A r ather small ridge (cr. fh.) is found betw een the two foramin a of the depression, w ith its one end coalescing imp ercep tib ly w ith the front wall of 'the depression, wh ile the other end terminates as 'a fnee process on the back wall. From the foramina aft the floor of ith e hypophyseal d epr ession (f h.), two somew hat divergent furrows run across the fmnt w ail. The wal'ls of the depression which , frontally , r ath er gently partition off the posterior part of the parasphenoid from theanterior cordate part, at th e hack d rop d own s teeply toward the t hin list of t he poster ior rim. On the sides they form sub-even, triangudar p rocesses (prl. d.), transec ted by a deep notch. From underneath these, branch off lamell ar lateral proces­ ses of the dors al side, as mentioned here above. The present description has been based on the shape of the parasphe­ noid in Brachythoraci incertae fam iLiae and of the specimen referable to Dinichthys cf . pu stulosus . As figur ed in the d rawi:ng by Dun k le and Bun­ gar t, It alsoagrees with the st ructure of the parasphenoid in Dinichthy s 1erreHi and H eintzic hthys gouldii. Though the bad stete of p reservataon is a hindr ance in the way of ob­ servations, yet it would seem thaJt' the parasphenoid belonging to a new r epresentative of Brachythoraci is of a somewh at different structure. It sh ow s a similar curve in the anterior part; vencrally i t also is provided with a longitudinal crest t hough not so stron gly developed as in Dinich­ thys cf. pustolosue. Further , it probably lacked the t ransverse gr oove as well as ith e bridge and foramina piercing the element art m id-line. In rts dorsal region there probably ex is ted a single, blind, funnel-lik e p it poorly w alle d in..

DISCUSSION

In its shape th e parasphenoid of the Brachyithoracl di ff e rs con sider­ ably from the correspondmg element in Kujdanoviaspis buczaczie nsis where it is a Hat r hornbiforrn plat e, strewn w ith minute t ubercles and p ierced at mid-poirrt by the hypophysial Ior amen. In this representative of the Arctolepida the elem ent here considered ds in rio w ay relat ed to th e internad carotids w h ich p enetrat e t he sku ll later ally from the para­ sphenoid. The shape of the parasphenoid in Buctumosteu s osseu s is an interrne­ diata form between th at in Kujda novi aspis and in such r ep resentatives ON T HE P ARASPHENO ID OF THE BR ACH YTHORACI 107 of Brachy thor acras Dinichthys and Heintzichthys . It is also 'a rhom biform plate strewn with tubercles arranged along a somewhat radial pattern. On the other hand, instead of the single hypophysial forarnen, two fora­ mina w ,HI be here encountered ad mild-point, partntioned bya bone la­ mella. Beh ind 'thes e foramina there is 'a transverse groove w hich, in the opiri ion of E. S. Hills, unites wdthfhe blind sac-Iik e pits and commun i­ cates on each side with the canal w hich opens on the late r al bone margin. E. S. Hills favours the supposition that these two foramina are the hypo­ physia l canals, in spite of his OW!Il assertion that he knows of no case in which h ypophysis, eit her in em bryonic or adult stages, w as paired, That au thor also supposes that the lack of lamina in the hypophysial foramen in Kujdanoviaspis may have resulted from t he bad state of preservation of the specimen (1946, p. 218 : "...al though in a poorly preserved s peci m en it is qurte possible that rthe duplicate nature of this canal would not be obser vable ". This interpret ation m ay be doubted not only owing to the unpaired nature of the hypophysis but also the str ucture of the par asphenoid in Din ichthys and Heintzichthys. As may be inferred fr om the precedin g descr ipt ion , the paired foramen is in these forms situated wirtJhin the transverse groove , which testifies to the m utual association of these two structu res . Further, in between the groove fooamimlaa bone-bridge is encoun tered, which must h av e served as protection of some connecting structur e. It seems rather improbable forrthe glandular ducts. to have developed a IIlJetwJOI'Ik of this kind. It mightraoher be jnfe rr ed th,at inter­ n ai carotids, connected by anastomosy running und ern eath t he bridge, passed through the pained foramen. As mentioned b y Hills and as is to be seen from his drawings, the parasphenoid ,in Buchanosteus osseus has been damaged anthe region of tihe ·"bHlnd sac-like pitsvand canals lead ing to the transve rse groove. :Lt would appear that this m ight have led to w ant of exactitude in reconstruction and that the "bEnd sac-like pits" may be p acts of canals opening on the lateral m argin of the parasphenoid. Yet on the roth er hand, the absence oif any grooves or canals whatever, leading to the h yp ophysial forarnen in the Kujdanoviaspis buczacziensiJ bars all other in terpretat ions and suggests incorrectness in Hills' sup ­ posit ion that ,the had state of p reserv ation may 'be r esponsible for the ab­ sence of t he bone lamella. It is the probable absence of paired foramina, as well as that of the ,transverse groove on the parasphenoid of the represent ative of the new genus of Brachythoraci that p rohibirts full centairrty in confirmmg the correctness of interpretatacn in the case of these for amina, More con­ stancy is r1Jo b e expected ill vessels of such significa nce, though in the case of r ays, for instance, they vary t heie behaviour, en tering in to the 108 JUl. !.;" KULCZYCKJ sk ull cavity either indep enden tly by means of paired foramina, or again by means of a single one. As abo ve cited, the form her e considered probably approaches genus Phoiu iost eu n, which, with resp ect to some of i1ts characters, is more pri­ mrtive th an the Cocco steus and represents an independent line of evolutio n of the Brachythoraci. There isa chan ce, therefore, th at a somewhat dif­ ferent course was taken by its internal car otids which penetrated the skull laterally in relation to the parasphenoid, as in the case of K u jd a­ n oviaspis buczac ziensis, while th e transverse anastomosy, if it did exist, might have run very superficially, either an the ventral or on the-dorsal side, without leaving signs of its course an the p arasphenoid. In conformity with the above, the parasphenoid in Buchanosteus osseusand in Brachythoraci would not be provided with the hypophysial Iorarnen presen t .iri K u jdanoviaspis buczacziensis. Since the presence of 1he hypophysial Ioramen is a prim ary character, while its absence a se­ condary one, we mu st recogruise when accep ting the a bove rirnte rpr etation . that: t he parasphenoid in K ujdanoviaspis buczacziensis r epresents a "more primitive stage than that encoun tered in Buchanosteus osseus an d t he Brachythoraci. Similarity in structure of the p arasphenoid, r eferable to Buciumosteus and that of Brachythoraei, is one more proof testifying to the close re­ lationship between the latter forms and Arctolepida. As mentioned above, the presence of the hypophysial foramen on the pacasphenoid of Kujdanoviaspis buczaczi ensis implies small probability to the suggestion that simplicity of str uctu re in this elem en t is referable to second ary simplification. In this connection H is to be inferred that the evolution of the par asphenoid within both, Arctolepida and Brachy­ thoraci, had taken place independently from that within other s tooks of fishes. The resultlngcharacteristic shape of the bone persisted in Brachy­ thoraci zo tJhe end of their existence. In asmuch as the relationship be­ tween other representatives of the Arthrodires with those under conside­ ration, has 110't been adequately clar ified and t alking imto accoun t that no decisive evidence has b een supplied showing some stage of reduction of Ithe parasphenoid in Arthrodires (it must be a dded that not long ago the Arctolepida and Brachythoraci were regarded as lacking this skeletal element)- suggestions concermimg the dtsirutegretion and reduction of the parasphenoid in the Arthrodires does not seem justifiable in the light of data now available. The external appear ance of the parasphenoid of Brachythorac i re­ sembles that elemen t in Porolepidae as described and figured by Gross (1936) an d Jarvik (1954). In these fishes the parasphenoid is also rel at i­ vely shor t an d broad, w ith a: di stinct anter.ior part separ ated from th e 0 1\' TilE P ARASPlI E:'-I OID OF T H E 13RACH YTIlOlnC I 109' poster ior one by what Jarvik (1954) has called prespiracular groove. In Gly ptolepis there is a hypophysial Ioramen in the mid-line of this groove. At fi rst sight these morphological d etails m ay appear identical with the transverse groove and ,the pai red Ioramen associated with it, as descr ib­ ed above. Yet , in Porolepid ae, the presp ir acular groove lies superficially on ly covered with a mucous Un.ing of the oral cavity. T his is testified by the presence of ornamentation. Concrarywise , the ornamentation of the paras phenoi d of Brachy thor aci, persisting jn Dinichthy s cf. pustulosus, occurs in t h e anterior co rdate part only, w hile it is lacking in the groove. This condition and the presence of the "bone-br idge", whose only pur­ p ose could h av e been .the prot ec tion of under lying structures, both testify that the groove must have been associated wuth cer tain anatomic struc­ tures by the present writer believed to 'have been the internal crotids . It would thus seem that the prespiracular groove and foramen on the parasphenoid of Porolepidae cannot b e homologised with the groove and foramin a on the parasphenoid of Brachy thoraci. By its external appear ance the parasphenoid of Brachythoraci also resem bles the anterior - basal ele ment of the Acanthodians, described by Watson (1937), this bein g stressed by Dunk le and Bungert (1946). Still ­ in opposition to the parasphenoid of Brachy thoraci, which is a typical dermal bon e as testified by its 'h is tological structure and also by the presence of ornamentation persisting on the pa rasphenoid of Dinich th y., cf . pu stulosus - the anterior-basal element in the Aoamthcdians shows perdohcndrial ossification only. Neither foramina nor grooves for th e in ternal carotids Me to be found on .the ventral side of the anterior­ basale of the Acant hod ians. Th e forame:n of the hypophysial canal' is situated in the centre of this element.F ur therm ore, 'th e anterior-basal elemen t differs from the parasphenoid of Brachy thonaci by a n umber of other, less significant characters. The lack in the available specimen of Brachy thoraci of that p ar t of endocranium adjacent to the parasphenoid, makes impo:ssible a close an alysis of this dermal element in 'its relation to the endocranium, or of the relation of its lateral processes to S'imilarr structures on the anter ior basal of the Acanthodians and the ascending processes of the Crossopte­ rygians and .Actinopter ygiens,

Paleozoological Laboratory of the PoUsh Academ y of Sci en ces Wa r szawa , N ov ember 1955, 110 J ULl A:-I KULCZYCKI

REFERENCES

DUNKLE D. H . &: BUNGART P . A. 1946. The Arntero-Supragnathal of Gor go nich­ thys. Amer . Mus. N ov it., 1316. Ne w Yonk, GROSS W. 1936. Be itrage zur Osteologie baltischer umd r heimdscher D evon-Cros­ sopterygier. P alaeont. Ztschr., 18, 129 - 155. Ber ldn. H ILLs E . S. 1936. On certain endocranial structures iJn Coccosteus, Geot. M ag., 73, 21'3 - 226. London. ·J ARVIK E. 1954. On the vdsceral sk eleton iIn Eustheno p teron with a discussion of the panasphenodd an d patatoquadrate in fishes. Kung!. Sv . Vetensk. H andL, Ser. F , 5, 1. 1 - 104. S tockholm. S TENS IO E. A. 1942. On .the snout of Anthrodires. Ibidem , S er. 3, 20, 3, 1 - 32. 1945. On t he heads of certadn Arthrodires. 11. On mheoranirum an d cerv ical jo iJnJt of the DoiliJchothor aci (Acan thasp ida). I bi dem., S er. 3, 22. 2, 1 - 70. WATSON D. M . S . 1937. 'I'he Acarrthodiari Fdshe s, P hilos. Trans. Roy. Soc. L on don, S er. B , 228, 549, 49-146. Lo ndon .

J ULl A:-i KULCZ YCKI

o P ARASFENOIDZIE RYBBRACHYTHORACI

Streszczeni e

Notatk a zawiera opts parasphenoideum B rachy tho ra ci z gonne go dewonu K ie1c o raz IPO,rOwnlarne z o dp ow dedn drn elementem u Acanthaspida vi, ndektorych Crosso­ p te r ygdi (Porolepidae), jaik rownoiez zeeutero-basaue Acanbhodui.

OB JASNJENIA DO IU 1STl L\CJ I

PI. I A antero-bazadny element Acaruthodid (wg W:wtsona), B p arasfenodd B uchalw­ st eus osseus tws Hiblsa),C parasfenold D i ni ch thys t errelli (wg Dunkle'a i Bungarta), D parasfenodd H ei ntzichthys gouldii (wg Dunkle'a do Bungantaj, E p arasfen oid Glyptolepis, F podl uzny przekro] przez parasfenodd B rachythoraci, G p arasfenoid D inichthy s cf. p ust ulosus, H parasfenoid Brach y th or a ci incertae familiae, A , B, E, G powierzchnia b rzuszna. C, D, H powierzchruia grzbietowa . a. tub. pole ipokryrte guzkami, bs. slepe wozkowate zagtebiende, ca. kanal dla anaetomozy tetndczej, c. nci. lkan a l d la anteria carotis lilnt erna, c. hypo k ana l .p rzy­ sadkow a ty . c r. fh. g rzeb ien zag lebiema .pr zysadkow ego, cr. m. grzeb ien sr odkowy, f h. zaglebiende przysadkowe, gr. p sp , r owek .preiip irakulcmmy, p . m ostek k osmy, p. a. przednia sercowata m;~s c, pp tyln a-czes « p arasfe no idu, pr], d. grzb letowy w yrostek boczny , pr ]. v. b rzuszny wyrostek boczmy, se. m aly r owek z tylu 'kostnego mostka. s m . r ow ek srodkowy, s. t r . rowekpoprze czny, t. gu zek. ON THE PARASP HENOlD OF TH E BRACHYTHOR.·\C r 11:

PI. If F ig . 1 - 2. Paras fenond Brachythoraci n. gen.; 1 - s trona grzbietowa , 2 - strona br zuszna; X 1,5. F dg, 3. P arastenosd D inichthys d . pust 1~ l os u s., s trona brzuszna; wielJk. nad. F'ig ; 4 - 7. P ar aefenoid Brachythoraci in certae familiae; 4 - s trona brzuszn a, 5 - strona grzbieto wa, 7 - iSUroI11JCl boczma X 1,5; 6 - sr.odJkOWla cz~sc poprzecznego r owka w r zu cie ukosnyrn, iskaeu jacym o twor iPOd .k ostnyrn m ostkiem d la anasto- m ozy .t etn iczej ; X 3. '

IOnl1/1H KYJIb4Y1l1KYI

o IT APAC

P eSIOMe

Ilpennaraexaa CTaTbH CO.IlepmHT OI1HCaHHe napacrpeaoana .pbI6 Brachy thora ci H3 sepxaero ne sona Hensu H ero op aaae ane C COOT3eTCTBYIOl..l\HlVI 3JIeMeHTOM npencra ­ RWr eJJ eti Acannhaepida 'H HeKOTOpbIX Crossopterygit (P orolepid a e), a r arox e C anr epo­ 6a3 aJJ bHbIM 3JleMeHTOM Acamtho dhi.

ACTA PALAEONTOLOGICA POLO~IC A . VOL. I J . KULCZYCKI. PLo 1

A B c o E

Ib. c.aci. crib .

sm. G H

ea cl.

,, - \ t \ I , I \ I I I , \ I , I , / I ',. / I , -, / , / / ...... ' '- / ... " ' ~"'*"- "" -,__ -", X __- - ' J " p pp se. prl.v. crin.

A anterior-basal element of the Acanthodians (after Watson). B paraephenodd of Buehanosteus osseus (after HilllliS), C paraephenodd of D i ni ehthy s terreHi (after Dunkle & Bungart), D paraephenoid of Heintziehthys gouldii (adlter Dunkle & Bun­ gara), E pa'I'l3Ejpheno1d of Glypt ol epi s, F longatudinal se ction of the ,paJl'aspheno:d oil Brachythoraci, G parasphenoid of D i nk ht hys cf. pusrulosus, H .parasphenoid of the Brachythoraci incertae familiae, A, B, E, G vent-rail aspect, C, D,H dorsal aspect. a. tub, tuberculaned area, bs. blind sac-like P'$, cc. OO'IlJaIl for th e aetertal ana­ St'OffiOSUIS, e. aei c anall foil' the arteria oarotds l iJ nt ~l'I1J9. , c. hypo hypophysial camat, er, fh. crest olf I1Jhe hypophysiral dqpressiom, cr. m. m ed:17!TI crest, fh. hypophysial depressdon , gr. psp, IPI1es!P':'ra,cu},aif groove, p. bone-bridge. pa. anterior cordate P aJI'.t, pp,posltelf:OII" part of the parasphenold, pr], d. dorsal la'tenad process, pr], v. ven tral lateral pmee.;-s. se. sma[[ groove b ehind ,the bone-bridge, sm. median groove, s. tr. transverse groove, t. tubercle. .\ CT.\ 1',\\..\ EO:-lTU I.Oli\ C,\ 1'0 1.0 :\I C.\ . \ ·( ll .. \ .I. I,U I. CZ\ CK I . PLo I1

2

3

4 5

6 Fig. 1-2. Parasphen oid of Brach ythoraci n. gen .; I-dorsal, 2- ventral aspects; X 1,5. l ·ig. 3. Paraephenoid of Dinichthys cf. pllstu loSHs; verul.ral as pec t ; nat. size. Fig. 4-7 . P araspben oid of Brachythor ac i incertae familiae; 4-venLrall aspect, 5-do r ­ sal aspect, 7 - lateral aspect, X 1,5; 6 - mid-pant of the transverse groove in obliqu e pr ojection s howing the a nastomotic forarnen under th e bone­ bridge, X 3.